Quality Appraisals

The quality appraisals are presented in Table S4 and Table S5 (Supplementary Material; available online only) of the supplement. Overall, eligible studies showed a moderate to high risk of bias. The two tools used provided similar results.

Subjects Characteristics and Outcomes

A summary of each study is presented in Table 1. Based on the limited available data scattered throughout the eligible studies, ^{Reference Nozaki, Hurl and Shinozawa37–Reference Yanagisawa, Morita and Nakajima40} this study estimates that a minimum of 8,550 individuals were exposed during the sarin gas attacks that struck Japan and Syria, of which 1,333 casualties and 11 deaths were confirmed medical cases managed on the day the attacks occurred (Supplementary Material, Table S6 and Table S7; available online only). The 1,333 casualties represented confirmed chemical poisoning cases in acute settings and were considered in this study as the number of included patients.

Table 1. Summary of Included Studies

Abbreviations: CPR, cardiopulmonary resuscitation; CWA, chemical warfare agents; ED, emergency departments; M, Matsumoto; T, Tokyo.

* Unawareness of a CWA attack.

∫ – Design deduced from the paper as authors did not specify their design.

£ – The source of data is deduced from the paper as it was not provided by the authors.

† – Measurement not substantiated in the literature.

‡ – No biostatistics plan and analysis.

! – This represents the minimum number of patients managed by medical authorities over the years above the numbers treated in acute settings and reported in this paper.

€ – Secondary exposures confirmed by authors (ie, expansion of the contamination zone due to contaminated carriers [casualty/vehicle]).

§ – Signs of secondary exposures (ie, issues with PPE and decontamination capabilities, health care staff, and other rescuers becoming sick or absence of specialized capabilities).

¥ – Visual Analogue Scale Grade (No information confirmed – Absence of information about the topic/category confirmed; Not reported – Uncertainty as to whether the authors might or might not have analyzed this topic/category; Partially – little information available; Detail(s) provided – Disclosure of the information).

Overview of the Populations Treated in Acute Settings

Nozaki, et al was the only study to provide a complete basic breakdown of the affected population (n = 15 medical staff members; 13 males, two females, all Japanese, ages ranging from 25 to 51 years old). ^{Reference Nozaki, Hurl and Shinozawa37} In the Okumura, et al study, 640 patients were treated but the authors only provided a partial breakdown (395 males, five pregnant females; aged eight to 65 years old). ^{Reference Okumura, Takasu and Ishimatsu38} No information was provided on the remaining 240 individuals poisoned. ^{Reference Okumura, Takasu and Ishimatsu38} In their study, Yanagisawa, et al reported that the 1994 Matsumoto attack resulted in a total of seven fatalities and 272 casualties treated the day of the attack (264 patients; eight rescuers). ^{Reference Yanagisawa, Morita and Nakajima40} In the 1995 Tokyo attack, the same authors reported four dead and 920 survivors treated the day of the attack: (1) St. Luke’s Hospital: 750 (one dead; 749 affected individuals, 639 patients and 110 medical staff members); (2) Keio University: one dead, 85 patients, 15 medical members; (3) Teishin Hospital: 32 patients and 39 rescuers; and (4) Tokyo Subway Station: two dead. ^{Reference Yanagisawa, Morita and Nakajima40} Age and gender were not reported. ^{Reference Yanagisawa, Morita and Nakajima40} Rosman, et al provided a casualty estimate (n = 130; 3% females, 97% males; of which 60% were children) based on their source of data (YouTube [Google, Inc.; San Bruno, California USA] social media footage analysis). ^{Reference Rosman, Eisenkraft and Milk39}

Medical Interventions During Casualty Extraction

None of the four papers ^{Reference Nozaki, Hurl and Shinozawa37–Reference Yanagisawa, Morita and Nakajima40} provided comprehensive details regarding treatments given to patients as a function of symptomatology during the medical extraction. In the Nozaki, et al study, no details were provided regarding the medical interventions performed on the 85 contaminated patients upon arrival at Keio University Hospital. In addition, their clinical presentation during the medical extraction from the chemical attack site in Tokyo and once admitted to hospital was not reported by the authors. ^{Reference Nozaki, Hurl and Shinozawa37} However, the authors did report some information on two patients’ respective health conditions, one convulsive and one in cardiac arrest, for which cardiopulmonary resuscitation (CPR) was performed upon the transfer to the emergency department (ED)/emergency room (ER). ^{Reference Nozaki, Hurl and Shinozawa37} Table 2 lists the available information related to the continuity of care provided by the Japanese medical centers at the patient’s admission.

Table 2. Listed Treatments Patients Received Once Admitted

Note: There was no indication on the use of oxygen found in these studies.

Abbreviations: PAM, pralidoxime iodide; 2-PAM, 2-pyridinealdoxime methiodide.

In Okumara, et al, one person performed CPR on a victim at the site of the chemical attack in Tokyo before getting poisoned by sarin herself. At her arrival at St. Luke’s ER with two other victims, that Samaritan also was in cardiac arrest. ^{Reference Okumura, Takasu and Ishimatsu38} Regarding medical extraction, the authors only reported the transportation performed by paramedics. ^{Reference Okumura, Takasu and Ishimatsu38} No information was provided on the medical interventions performed on 99 patients during their transport to hospital by first responders. ^{Reference Okumura, Takasu and Ishimatsu38} Similarly, the authors did not report on first aid performed by good Samaritans or health care staff for the remaining 541 rescued patients. ^{Reference Okumura, Takasu and Ishimatsu38}

Yanagisawa, et al reported that all cardiac arrest patients from Matsumoto (1994; n = 3) and Tokyo (1995; n = 5) were treated upon arrival to the ED, but no further detail was provided. ^{Reference Yanagisawa, Morita and Nakajima40}

In Rosman, et al, the authors listed treatments provided to patients in non-medical facilities: atropine, steroids, furosemide, supplemental oxygen (O₂), nasopharyngeal suctioning, bag valve ventilation, tracheal intubation, mechanical ventilation, and chest compression. ^{Reference Rosman, Eisenkraft and Milk39} They noted that standard monitoring equipment was not used to measure O₂ saturation, blood pressure, or cardiac electrical activity. ^{Reference Rosman, Eisenkraft and Milk39} As noted by the authors, all medication was administered intravenously with no evidence of autoinjector use. ^{Reference Rosman, Eisenkraft and Milk39} Moreover, they casted doubt on the authenticity of 66 out of 67 YouTube videos analyzed.

The two studies of the 1995 incidents that occurred in Japan ^{Reference Okumura, Takasu and Ishimatsu38,Reference Yanagisawa, Morita and Nakajima40} did not report whether the delivery procedures or special gestational care successfully preserved the life of the fetus/newborn or not. Similarly in the Damascus incident, Rosman, et al only mentioned that children accounted for 60% of the 130 casualties. ^{Reference Rosman, Eisenkraft and Milk39} Okumara, et al reported an eight-year-old victim as the youngest casualty treated at St. Luke’s Hospital, but no further clinical information was provided. ^{Reference Okumura, Takasu and Ishimatsu38} Likewise, no information was reported for specific populations.

Medical Interventions Due to Secondary Exposure in Rescuers and Medical Staff

Nozaki, et al was the only study ^{Reference Nozaki, Hurl and Shinozawa37} of the four ^{Reference Nozaki, Hurl and Shinozawa37–Reference Yanagisawa, Morita and Nakajima40} to have reported a medical response involving medical staff affected by managing patients contaminated by sarin, which led to a secondary exposure or the relocation of the contaminated zone to an unprepared location. The authors reported a six-hour wait before medical staff received confirmation that sarin gas was the cause of patient intoxications. ^{Reference Nozaki, Hurl and Shinozawa37} In the interim, medical staff provided medical care for an unknown exposure. ^{Reference Nozaki, Hurl and Shinozawa37} The authors briefly described some close-contact events that occurred between 15 clinicians and 85 contaminated patients. ^{Reference Nozaki, Hurl and Shinozawa37} Of these cases, only two were summarily described (the medical management of one convulsive and one cardiac arrest case). The authors also enumerated 14 of 15 reported total cases involving medical staff according to the following categories: four cardiac arrests, two intubations, three cases of contamination, four unspecified tasks, and one observation. Of these, six adult caregivers received atropine (0.5-1.0mg intramuscular); one caregiver received 2-PAM (500mg). ^{Reference Nozaki, Hurl and Shinozawa37} Other causative factors that led to a secondary exposure are covered in the Results section and the Supplementary Material (Body Text; available online only).

Medical Algorithms

Throughout the four studies, ^{Reference Nozaki, Hurl and Shinozawa37–Reference Yanagisawa, Morita and Nakajima40} there was no indication that specific chemical intoxication algorithms or clinical guidelines for patient management were used, except for triage. ^{Reference Nozaki, Hurl and Shinozawa37,Reference Rosman, Eisenkraft and Milk39} Medical authorities in Matsumoto ^{Reference Yanagisawa, Morita and Nakajima40} (1994) and medical staff at St. Luke’s Hospital (1995) used an algorithm ^{Reference Okumura, Takasu and Ishimatsu38} for patients’ triage (mild, moderate, and severe) and management, ^{Reference Okumura, Takasu and Ishimatsu38,Reference Yanagisawa, Morita and Nakajima40} even if the exact terms used varied. It should be noted that even though the two studies partly covered the same medical response at St. Luke’s Hospital, ^{Reference Okumura, Takasu and Ishimatsu38,Reference Yanagisawa, Morita and Nakajima40} their respective authors did not report precisely the same version of the triage score. No gold-standard reference related to that triage score was found in either study. ^{Reference Okumura, Takasu and Ishimatsu38,Reference Yanagisawa, Morita and Nakajima40} The definitions are shown in Table S8 of the Supplementary Material (available online only).

Immediate and Specialized Decontamination Capabilities

None of the four papers ^{Reference Nozaki, Hurl and Shinozawa37–Reference Yanagisawa, Morita and Nakajima40} reported whether or not immediate decontamination procedures were performed during the extraction process before their arrival at a specialized decontamination asset before admission to a medical facility, usually considered as a clean zone. None of the papers reported the existence of specialized assets capable of combining actions like continuing medical treatments, performing decontamination, and ensuring safety while wearing personal protective equipment (PPE). ^{Reference Nozaki, Hurl and Shinozawa37–Reference Yanagisawa, Morita and Nakajima40}

However, three papers ^{Reference Nozaki, Hurl and Shinozawa37–Reference Rosman, Eisenkraft and Milk39} out of the four ^{Reference Nozaki, Hurl and Shinozawa37–Reference Yanagisawa, Morita and Nakajima40} provided details on some of the decontamination means used on patients while at the medical facilities. Nozaki, et al reported: (1) ventilation of resuscitation rooms was ensured by opened doors and windows; and (2) contaminated belongings were placed in sealed vinyl bags. ^{Reference Nozaki, Hurl and Shinozawa37}

Okumara, et al summarized decontamination steps as the removal of contaminated clothing. ^{Reference Okumura, Takasu and Ishimatsu38} They also reported that patients were either showered or bathed depending on their state of consciousness, but provided no further detail. ^{Reference Okumura, Takasu and Ishimatsu38} In the case of Rosman, et al, their observation of procedures was reported as: (1) wash out with water, which included rubbing the casualty’s face and chest (25% of videos); and (2) full removal of clothing (10 out of 67 videos in which decontamination took place at medical facilities with no additional information provided). ^{Reference Rosman, Eisenkraft and Milk39}

Personal Protective Equipment

None of the four studies confirmed PPE was worn during the management of the chemical attacks. ^{Reference Nozaki, Hurl and Shinozawa37–Reference Yanagisawa, Morita and Nakajima40} Rather, while two studies made no mention of PPE for rescuers, the clinicians, and the patients, ^{Reference Nozaki, Hurl and Shinozawa37,Reference Okumura, Takasu and Ishimatsu38} the two remaining studies presented little information on their means of protection. ^{Reference Rosman, Eisenkraft and Milk39,Reference Yanagisawa, Morita and Nakajima40} The authors reported that health care professionals were not protected from contamination despite the suspicion of gas poisoning. ^{Reference Yanagisawa, Morita and Nakajima40} Regarding the Tokyo attack, the authors confirmed rescue staff did not use special PPE to protect themselves against the gas exposure. ^{Reference Yanagisawa, Morita and Nakajima40} They neither specified the members of the rescue teams nor if the medical staff used any PPE. ^{Reference Yanagisawa, Morita and Nakajima40} In Rosman, et al, it was reported that no PPE was worn other than the sporadic use of latex gloves and surgical masks by medical staff (10 out of 67 videos). ^{Reference Rosman, Eisenkraft and Milk39}

Other Causes of Secondary Exposures

As previously indicated, Nozaki, et al ^{Reference Nozaki, Hurl and Shinozawa37} was the only study of the four ^{Reference Nozaki, Hurl and Shinozawa37–Reference Yanagisawa, Morita and Nakajima40} that covered the medical management of new patients (ie, medical staff) due to secondary exposures induced by contaminated patients. For its part, Yanagisawa, et al identified failures in PPE capabilities as a cause of secondary exposure to rescuers for the Matsumoto and Tokyo chemical attacks. ^{Reference Yanagisawa, Morita and Nakajima40}

Meta-Analysis

Due to the paucity of studies and the heterogeneity of the data, no meta-analysis was performed.